FIELD OF THE INVENTION

[0001]
The present invention relates to a method and a device for power based radio resource management in wireless radio systems, such as wireless CDMA systems.
BACKGROUND OF THE INVENTION

[0002]
In wireless radio systems, such as the third generation (3G) system, radio resource management (RRM) is responsible for utilisation of the air interface resources. RRM is used in order to guarantee the socalled Quality of Service (QoS), to maintain the planned coverage area and to offer high capacity to the users. RRM can be divided into different functionalities, such as handover control, power control, admission control, load control and packet scheduling functionalities. These functions are required to guarantee the Quality of Service and to optimise the system data throughput with a mix of different bit rates, services and quality requirements.

[0003]
RRM algorithms can be based on the amount of hardware in the network or on the interference levels in the air interface. The case where the hardware limits the capacity before the air interface gets overloaded is called “hard blocking”. The case where the air interface load is estimated to be above the planned limit is called “soft blocking”. It has been shown that soft blocking based RRM is advantageous as it provides higher capacity than hard blocking based RRM. Therefore, the present invention is concerned with soft blocking based RRM.

[0004]
In case of utilising soft blocking based RRM, the air interface load needs to be measured. The estimation of the uplink load of the air interface can be based on the wideband received power level or on throughput. The present invention is engaged with load estimation based on wideband received power.

[0005]
The received power levels can be measured in the base station. Based on such measurements, the uplink load factor η can be obtained. The corresponding calculations are explained hereinafter with reference to FIG. 1.

[0006]
FIG. 1 shows a base station BS including an antenna 1. For simplicity it is assumed here that one BS equals one cell. The concept can however be extended to cover the case where one BS has several cells. The base station BS receives via the antenna 1 an owncell interference power I_{own }from all intracell users connected to the base station BS. Furthermore, the base station receives via the antenna 1 an othercell interference power I_{oth }from all intercell users that are utilizing the same carrier frequency but are connected to other cells than this own cell.

[0007]
Furthermore, the base station BS receives system noise with a system noise power P_{N }via the antenna 1 as well as from its own system components, i.e. system noise is at least partly inherent in a base station BS.

[0008]
The owncell interference power I_{own}, the othercell interference power I_{oth}, and the system noise power P_{N }represent the received wideband interference power, called total received power. This can be expressed by the following equation:
I _{tot} =I _{own} +I _{oth} +P _{N }

[0009]
The total received power is measured continuously by means of a measurement circuit 3A. The system noise power P_{N }can be measured by the base station by means of a measurement circuit 2. The system noise power P_{N }is commonly estimated at night, when the load is assumed to be small. Thus, the owncell interference power I_{own }and the othercell interference power I_{oth }are small as well. This results in
P_{N}≈I_{tot }

[0010]
The thus estimated noise power P_{N }is then used in an RRM controller 3 to perform RRM functionalities, such as load control and admission control.

[0011]
Unfortunately, this method cannot cope with system noise differences between day and night. Furthermore, this prior art method does not allow to determine the othertoown cell interference ratio i at all, namely the ratio of the otherall interference power I_{oth }to the ownall interference power I_{own}. Thus, rather conservative noise rise targets have to be used in load control. This degrades systemperformance.
SUMMARY OF THE INVENTION

[0012]
It is therefore an object of the present invention to improve the radio resource management.

[0013]
This object is achieved by a method for power based radio resource management in wireless radio systems comprising the steps of

 continuously measuring a total interference power I_{tot }received at a base station,
 continuously measuring an owncell interference power I_{own }of all intracell users connected to a predetermined cell, received at said base station,
 continuously estimating a system noise power P_{N }and/or continuously estimating an othertoown cell interference ratio i based on a number of consecutive measurements of said total interference power I_{tot }obtained by said total interference power I_{tot }measuring step and based on a number of consecutive measurements of said owncell interference power I_{own }obtained by said owncell interference power I_{own }measuring step, and
 performing at least one functionality of said radio resource management based on said estimation of said system noise power P_{N }and/or said othertoown cell interference ratio i.

[0018]
Furthermore, the above object is achieved by a device for power based radio resource management in wireless radio systems comprising:

 means for continuously measuring a total interference power I_{tot }received at a base station,
 means for continuously measuring an owncell interference power I_{own }of all intracell users connected to a predetermined cell, received at said base station,
 means for continuously estimating a system noise power P_{N }and/or continuously estimating an othertoown cell interference ratio i based on a number of consecutive measurements of said total interference power I_{tot }obtained by said total interference power I_{tot }measuring means and based on a number of consecutive measurements of said owncell interference power I_{own }obtained by said owncell interference power I_{own }measuring means, and
 means for performing at least one functionality of said radio resource management based on said estimation of said system noise power P_{N }and/or said othertoown cell interference ration i.

[0023]
The present invention improves the performance of RRM systems, in particular RRM functionalities such as admission control and load control. In power based radio resource management there are two important system parameters related to the uplink (reverse link), namely the othertoown cell interference ratio i and the system noise power P_{N}. Knowledge of these parameters is advantageous for certain RRM functionalities such as load control and admission control. These parameters are very useful for radio network planning and optimisation purposes. The present invention enables an online estimation of the othertoown cell interference ratio i and the system noise power P_{N}. In particular, as the othertoown cell interference ratio i and the system noise power P_{N }are time varying a robust online estimation of these parameters is desirable. Online knowledge of these parameters is useful for e.g. load estimation and identification of cells with interference problems, e.g. a high othertoown cell interference ratio i.

[0024]
Due to more exact estimates of system noise power P_{N}, less conservative noise rise targets can be used in load control, which in turn means that a higher capacity can be reached for that particular cell.

[0025]
The present invention further provides new possibilities to network planning and optimisation, since the interference situation in each cell can be monitored online. Cells with potential problems are easily detected and troubleshooting becomes easier.

[0026]
When using power based load control, accurate knowledge of system noise power P_{N }is desirable. Therefore, preferably, the uplink load factor η is continuously calculated as
$\eta =1\frac{{P}_{N}}{{I}_{\mathrm{tot}}}$
wherein I_{tot }is said estimated total interference power and P_{N }is said estimated system noise power.

[0027]
Alternatively the uplink noise rise NR is continuously calculated as
$\mathrm{NR}=\frac{{I}_{\mathrm{tot}}}{{P}_{N}}\text{\hspace{1em}}$
$\mathrm{or}\text{\hspace{1em}}\mathrm{as}\text{\hspace{1em}}$
$\mathrm{NR}=\frac{1}{1\eta}.$

[0028]
The system noise power P_{N }varies over time. For example, man made noise, e.g. from engines etc., is added to the system noise and can be considerably higher at rush hours than at night. Therefore, preferably, the system noise is estimated online, continuously, in order to allow a more accurate load factor /noise rise calculation. In particular the noise rise NR calculation shows that an accurate system noise power P_{N }estimation is advantageous as an error in the system noise power P_{N }yields an error in noise rise NR.

[0029]
Further advantageous developments are defined in the dependent claims.
BRIEF DESCRIPTION OF THE DRAWINGS

[0030]
In the following, the present invention will be described in greater detail based on preferred embodiments with reference to the accompanying drawings, in which:

[0031]
FIG. 1 shows a base station with a power based radio resource management system according to the prior art;

[0032]
FIG. 2 shows a base station with a power based radio resource management system according to an embodiment of the present invention comprising online estimation of system noise power P_{N }and othertoown cell interference ratio i;

[0033]
FIG. 3 shows a first embodiment of the online estimation of the system noise power P_{N }and the othertoown cell interference ratio i shown in FIG. 2;

[0034]
FIG. 4 shows a second embodiment of the online estimation of the system noise power P_{N }and the othertoown cell interference ratio i shown in FIG. 2;

[0035]
FIG. 5 shows a third embodiment of the online estimation of the system noise power P_{N }and the othertoown cell interference ratio i shown in FIG. 2;

[0036]
FIG. 6 shows a chart illustrating the performance of the system noise power P_{N }estimation; and

[0037]
FIG. 7 shows a chart illustrating the performance of the othertoown cell interference ratio i.
DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0038]
FIG. 1 has been explained with reference to the prior art in order to explain the background of the invention. Therefore, the above explanations with regard to FIG. 1 apply to the invention as well, as far as nothing else is described hereinafter.

[0039]
FIG. 2 shows a schematic diagram of a base station BS including an antenna 4 for communication between a radio network, such as a UMTS network utilizing wireless CDMA, and user equipment (not shown), e.g. mobile phones or any other mobile devices.

[0040]
Such a base station BS covers a certain area in order to establish a wireless connection between the base station BS and the user equipment(s) being located within this area. Such an area is defined as a cell.

[0041]
However, practically, there can be user equipment within a certain cell that is connected with the base station of another cell. Such user equipment causes interference with the signals that are intended for this specific base station BS. Furthermore, user equipment may interfere with this specific base station even though they are not located within the cell of this specific base station BS. These interferences can be regarded as othercell interference for all intercell user equipments that are connected to other cells as far as they are utilizing the same carrier frequency as this specific base station BS. The power of this othercell interference is defined as othercell interference power I_{oth}.

[0042]
It is noted that I_{own }and I_{oth }refer only to interference sources within the same frequency band as the base station BS in question. Interference from sources in other frequency bands may arise due to nonperfect filters. For example, UEs operating on another frequency band may have nonnegligible adjacent channel leakage into the frequency band of interest, thus causing interference. This interference is covered in the system noise term, P_{N}, and is one of the reasons for the timevarying nature of P_{N}.

[0043]
Furthermore, all the user equipments connected to this certain cell (hereinafter referred as owncell) cause an owncell interference. The owncell interference power that is received at the base station of this owncell (the power that is caused by all intracell users connected to the owncell) is defined as I_{own}. Note that the own cell interference is actually the useful part of the received power, carrying the transmitted user data from the user equipments (UEs).

[0044]
Furthermore, the base station BS receives background noise, in particular via the antenna 4, and a noise caused by the receiver section of the base station BS, namely the receiver noise. This background noise and the receiver noise as well as any other noise that by occur in such a radio system is defined as system noise with a system noise power P_{N}.

[0045]
The sum of the othercell interference power I_{oth}, the owncell interference power I_{own }as well as the system noise power P_{N }is hereinafter referred as the total uplink interference power I_{tot}.

[0046]
The total uplink interference power I_{tot }varies over time and is seen by the base station BS at every time instant n as:
I _{tot}−(n)=I _{own}(n)+I _{oth}(n)+P _{N} ^{−}(n)^{−}

[0047]
This total interference power I_{tot}(n) is continuously measured by a continuous total interference power I_{tot }measurement circuit that is comprised in the base station BS.

[0048]
Furthermore, the base station BS comprises a continuous owncell interference power I_{own }(n) measurement circuit 6 for continuously measuring the owncell interference power I_{own }(n).

[0049]
Both measurement circuits 5, 6 continuously provide consecutive measured values I_{tot }(n) and I_{own }(n), e.g. every 100 ms. The provided value should reflect the average value over the measurement period. Thus, n can be seen as every measurement reporting instance of the base station BS.

[0050]
The values of tot (n) and I_{own }(n) are provided to online estimation means 7 that are implemented by software and/or hardware, e.g. by an online estimation circuit. The online estimation means estimates based on the received values I_{tot }(n) and I_{own }(n) the system noise power P_{N }as well as an othertocell interference ratio i that is defined as:
$i\left(n\right)=\frac{{I}_{\mathrm{oth}}\left(n\right)}{{I}_{\mathrm{own}}\left(n\right)}$

[0051]
Thus, the total interference power I_{tot }can be written as
I _{tot}(n)=I _{own}(n)+i(n)·I _{own}(n)+P _{N} =I _{own}(n)·(1+i(n))+P _{N}((n)

[0052]
The online estimation means 7 assumes that the total interference power I_{tot }and the owncell interference power I_{own }are continuously measured by the base station BS, in particular by the measurements circuits 5, 6 and utilizes these measurements to estimate the othertoown cell interference ratio i and the system noise power P_{N}. FIG. 3 illustrates the online estimation circuit 7 of FIG. 2 which is denoted as 307 in FIG. 3. A number of consecutive measurements of the total interference power I_{tot }and the owncell interference power I_{own }is carried out. Thus a system of equations is set up according to:
$\left[\begin{array}{c}{I}_{\mathrm{tot}}\left(n\right)\\ {I}_{\mathrm{tot}}\left(n+1\right)\\ \vdots \\ {I}_{\mathrm{tot}}\left(n+N\right)\end{array}\right]=\left[\begin{array}{c}{I}_{\mathrm{own}}\left(n\right)\xb7\left(1+i\right)+{P}_{N}\\ {I}_{\mathrm{own}}\left(n+1\right)\xb7\left(1+i\right)+{P}_{N}\\ \vdots \\ {I}_{\mathrm{own}}\left(n+N\right)\xb7\left(1+i\right)+{P}_{N}\end{array}\right]\text{}\text{\hspace{1em}}=\left[\begin{array}{cc}{I}_{\mathrm{own}}\left(n\right)& 1\\ {I}_{\mathrm{own}}\left(n+1\right)& 1\\ \vdots & \vdots \\ {I}_{\mathrm{own}}\left(n+N\right)& 1\end{array}\right]\xb7\left[\begin{array}{c}\left(1+i\right)\\ {P}_{N}\end{array}\right]$

[0053]
This set up of a system of equations is performed by set up system of equation means 308, e.g. by software implementation or by hardware implementation, e.g. by a corresponding circuit.

[0054]
According to a first embodiment of the online estimation means 7, namely as shown in FIG. 3 as online estimation circuit 307, it is assumed that the othertoown cell interference ratio i and the system noise power P_{N }are fairly constant over the sequence N (independent of n). Thus, the equation system can be solved by means of minimum mean square error method (MSE) that is implemented in equation system solving means 309, either by hardware or by software implementation. Thus, estimates of the othertoown cell interference ratio i and the system noise power P_{N }are found.

[0055]
Referring back to FIG. 2 these estimates of the othertoown cell interference ratio i and the system noise power P_{N }are provided to a radio resource management unit 8 that in turn is realized by hardware and/or software implementation. This radio resource management unit 8 performs the functionalities of the radio resource management based on the received estimates of the received system noise power P_{N }and the othertoown cell interference ratio i. These functionalities are e.g. load control, admission control, packet scheduling, power control, handover control, load estimation and/or identification of cells with interference problems.

[0056]
FIG. 4 shows a further embodiment of the online estimation means 7 shown in FIG. 2 being denoted with 407 in FIG. 4. Online estimation means 407 corresponds mainly to online estimation means 307 and thus comprises set up system of equation means 408 and equation system solving means 409. However, the estimates of the system noise power P_{N }and the othertoown cell interference ratio i are averaged over time in order to get more stable results. Therefore, respective average means 410 for the system noise power estimate P_{N }are provided in order to yield an averaged system noise power value {overscore (P_{N})} and average means 411 for the othertoown cell interference ratio i estimate are provided in order to yield an averaged othertoown cell interference ratio value {overscore (i)}.

[0057]
FIG. 5 shows a further embodiment of the online estimation means 7 of FIG. 2 being denoted as online estimation means 507 in FIG. 5. Online estimation means 507 is designed in order to provide even better estimates of the signal noise power P_{N }by using averaged values of the received powers I_{tot }(n) and I_{own }(n) over a predetermined period of time, e.g. 10 s. Further, a corresponding average value of the estimated othertoown cell interference ratio i is provided as well. In order to achieve this aim, online estimation circuit 507 comprises not only set up system of equation means 508 and equation system solving means 509 which correspond to means 308, 408 and 309, 409 respectively, but also comprises average means 510 for the continuously measured total interference power values I_{tot }(n) and owncell interference power I_{own }(n) provided by the measurement circuits 5, 6, respectively. Hence, averaged values for the total interference power and the owncell interference power are provided as {overscore (I_{tot})} and {overscore (I_{own})}.

[0058]
The measured values I_{tot}(n) and I_{own}(n) are provided to set up system of equation means 508 which in turn provides its results to equation system solving means 509. Equation system solving means 509 calculates the othertoown cell interference ratio namely î as described above.

[0059]
The estimated value for the othertoown cell interference ratio î is averaged by average means 512 in order to generate an averaged estimated value {overscore (î)} for the othertoown cell interference ratio.

[0060]
Based on this averaged estimated othertoown cell interference ratio {overscore (î)} and the averaged received powers {overscore (I_{tot})} and {overscore (I_{own})} an improved estimate for the system noise power {circumflex over (P)}_{N }can be calculated by calculation means 513 as:
{circumflex over (P)} _{N} ={overscore (I _{ tot } )}− {overscore (I _{ own } )}(1+{overscore (î)})
wherein {overscore (I_{tot})} is the average I_{tot}, {overscore (I_{own})} is the average I_{own }and {overscore (î)} is the average estimated i over the same time period. The reason why this estimate becomes better is that the error in the estimated i is scaled down, since {overscore (I_{own})} is always smaller than {overscore (I_{tot})}. Moreover, the fast fluctuating nature of i is averaged out, which gives a more stable performance.

[0061]
As a result better estimates {circumflex over (P)}_{N }and {circumflex over ({overscore (i)})} are achieved for the system noise power and the othertoown cell interference ratio i based on the averaged values {overscore (I_{tot})} and {overscore (I_{own})} as well as {circumflex over ({overscore (i)})}.

[0062]
All the above mentioned means comprised in the base station BS and in particular the online estimation means 7, 307, 407 and 507 can be implemented by means of hardware and/or software. In particular, as the above described estimation is not very demanding in terms of computing power and realtime requirements it can be easily implemented in software even though hardware implementation is possible as well.

[0063]
The performance of the estimation is dependent on the number of measurement samples N used in the minimum mean square error estimation. Even though N may be set arbitrarily, a reasonable value is in the range between 5 and 10.

[0064]
The performance of the above described system noise power and othertoown cell interference ratio i estimations have been tested using simulated data from a dynamic WCDMA network simulator. The results are shown in FIGS. 6 and 7. Even though the othertoown cell interference ratio is slightly underestimated, the system noise power estimation is close to the true level.

[0065]
FIG. 6 shows the performance of the system noise power estimation P_{N}. Each value plotted is derived using the average of all previous power measurements. The system noise power estimation converges to about −100.74 dBm, while the true system noise power was −100.9 dBm (constant), i.e. the noise power was overestimated by 0.16 dB. The uplink average noise rise in this simulation was 2.3 dB. N=10 consecutive measurements were used for every new estimate of the othertoown cell interference ratio i.

[0066]
FIG. 7 shows the othertoown cell interference ratio i estimation. N=10 consecutive measurements were used for every new estimation of the othertoown cell interference ratio and the result is filtered with an IIR filter with forgetting factor alpha=0.1. The average estimated othertoown cell interference ratio i was 0.32 whereas the true value was 0.43.

[0067]
It is noted that the present invention is not restricted to the preferred embodiments described above. In particular, the above described estimations can be performed in the radio network controller (the equipment in e.g. a radio network subsystem for controlling the use and the integrity of the radio resources) as well. Thus the estimation have not necessarily to be performed in the base station. Both, the base station as well as the radio network controller comprise a radio resource management functional part which is suitable to implement the above described estimations. Thus, the preferred embodiments may vary within the scope of the attached claims.